Solid-state transformers are a beautiful re-imagining of power distribution sub-stations. They could become economically appealing around 2030, especially in renewable- and AI-heavy grids. This 15-page report explores SST technology, quantifying costs, efficiency, challenges and implications (in grids, and for MOSFET/SiC demand).
A transformer is used every time voltage changes in the grid. The global transformer market is worth $65bn pa, but power grids are currently gripped by a transformer shortage. So will this open the door to new and improved technologies that can change voltage levels across the grid?
The functioning of conventional transformers is re-capped on pages 2-3.
Notable imperfections of conventional transformers include unidirectionality, fixed voltage ratios, highly customized designs, which therefore have long lead-times, and the inherent replication of power quality problems (harmonics, transients) between the input side and the output side, as spelled out on pages 4-5.
A solid-state transformer (SST) is envisaged to address the imperfections of conventional transformers, by using solid-state devices (SiC MOSFETs), link capacitors and a high-frequency transformer with lower losses. We explain each stage of how an SST works on pages 6-9.
The key debate is cost. SSTs are 5-10x more costly than conventional distribution transformers. However, comparing a conventional transformer with an SST is a bit like comparing an exercise bike with an entire gym. If you value all of the functionality in the SST, based on the prices of power electronic components, SSTs come into the money before 2030, as quantified on pages 10-13.
Notable companies in SSTs, and recent capital raises, are discussed on page 14.
One controversy that we wanted to address in the report is whether SSTs will excite the market for power MOSFETs or for SiC materials. We found less support for this idea than we had hoped, for the reasons on page 15.